"Patients with severe depression benefit as much from psychological therapy as they do from pills," says the Mail Online, reporting on a study comparing two commonly-used treatments.
The research compared modern antidepressants such as paroxetine, citalopram and fluoxetine, with cognitive behavioural therapy (CBT) – a type of talking therapy aimed at helping people change unhelpful ways of thinking and behaving.
It found both treatments worked more or less equally well for the initial treatment of people with moderate to severe depression. However, the study did not say if either treatment was effective at preventing depression from returning at some point in the future or who responds best to which type of treatment.
The authors of the review, which included 11 studies involving 1,511 patients, concluded that people should be offered a choice of treatment.
NICE guidelines recommend that people in the UK with moderate to severe depression should be offered a combination of an antidepressant and a talking therapy such as CBT or interpersonal therapy.
The study was carried out by researchers from the University of North Carolina, RTI International and Danube University, and was funded by the Agency for Healthcare Research and Quality.
The Mail Online reported the study reasonably well, although its suggestion that therapy could replace antidepressants was not borne out by the research.
Also, the headline used the phrase "happy pills" to describe antidepressants. Some people find this description offensive, as antidepressants are a treatment for a serious mental health disorder, which often have significant side effects, not an instant fix to make people happy.
The researchers carried out a systematic review and meta-analysis of randomised control trials (RCT). A systematic review is the best way to sum up the evidence about a topic, but it is only as good as the individual studies that go into it.
Researchers looked for RCTs that compared modern antidepressants with CBT as the first treatment for adults with moderate to severe depression.
They pooled the results to get an overall answer about how the treatments compared.
One difficulty in researching psychological therapies is that often similar therapies have different names, and therapies with the same name may vary, depending on the therapist. The researchers used a wide definition of CBT, which included problem-solving therapy and rational emotive therapy, as well as CBT.
Another difficulty is that people often drop out of trials of mental health treatments. The researchers decided to assume that everyone who dropped out, whatever their treatment, did not respond to treatment or get better. This may under-estimate the effects of the treatments, but unless drop-out rates are very different between treatments, the results should equal out.
The researchers assessed each study for problems that could unfairly influence the results. Finally, they checked their figures using statistical techniques to see whether including or excluding certain trials at high risk of bias significantly changed the overall results.
The review included 11 studies, with a total of 1,511 patients. It found that people treated with antidepressants and people treated with CBT were equally likely to respond to treatment (risk ratio (RR) for antidepressants 0.91, 95% confidence interval [CI] 0.77 to 1.07) and to get better (RR for antidepressants 0.98, 95% CI 0.73 to 1.32).
They had similar improvements on a questionnaire designed to measure symptoms of depression.
More people taking antidepressants dropped out of studies because of side effects from treatment, but the numbers were small enough that this could have been a coincidence.
The researchers found no difference in results between people taking antidepressants alone, and those treated with antidepressants plus CBT. They didn't find any studies that compared CBT alone versus CBT plus antidepressants.
The researchers said their results should be interpreted "cautiously" because of the overall quality of the evidence.
However, they conclude: "Given that the benefits of second generation antidepressants and cognitive behavioural therapy do not seem to differ significantly … and that primary care patients may have personal preferences … both treatments should be made accessible, either alone or in combination."
Previous research has shown that both second generation antidepressants and CBT can be helpful for people with depression. This study found that they seem to work about as well as each other.
The study has many strengths, including the fact it is a systematic review, and includes information from RCTs involving more than 1,500 people. However, the studies don't give us much information about adverse effects of treatments, or who responds best to which type of treatment. This is important, because what works for one person may not work as well for another.
Some doctors think people with severe depression need to be treated with antidepressants before they are well enough to engage with CBT. Some people have a strong preference for therapy rather than tablets, or vice versa. Many doctors also think the two treatments work best in combination, especially for those with more severe depression.
There are some limitations to this review, including the small sample sizes in each included study. Also, three of the studies included some people in the CBT group who were also taking antidepressant therapy which reduces the reliability of the findings.
Additionally, the review used data from the studies on outcomes after a period of 12 to 24 weeks. It did not say which treatment was likely to be more effective over the long term.
While this study is reassuring, guidelines in the UK already recommend both antidepressants and talking therapies, with talking therapies suggested first for milder forms of depression.
It would be helpful to have more studies looking at which treatments are best for which people – for example, whether women or men respond differently to the different types of treatment, or people of different ages or with different types of depression. This would help GPs select the best treatment for an individual patient.
In the meantime, the authors' suggestion that both should be offered so patients can choose which they prefer, seems sensible.